In the precision parts manufacturing industry, accurate dimensional measurement of complex internal geometries and microstructures is a critical factor determining product quality and reliability. However, conventional Coordinate Measuring Machines (C...
In the precision parts manufacturing industry, accurate dimensional measurement of complex internal geometries and microstructures is a critical factor determining product quality and reliability. However, conventional Coordinate Measuring Machines (CMMs) face fundamental limitations in measuring narrow internal features due to the physical size of probes and approach angle restrictions. Non-contact optical measurement methods also encounter challenges in ensuring data reliability due to shadow areas caused by complex curvatures and surface reflection characteristics. To overcome these physical limitations, this study proposes an indirect measurement system utilizing Polyvinyl Siloxane (PVS) impression material. The objective is to systematize the PVS-based replication technique as an independent, statistically validated measurement system and to provide objective evidence for its industrial applicability. For this purpose, experiments were conducted on specimens containing dovetail slots and complex geometries that are highly difficult to measure directly. A precise replication and slicing process was performed using PVS material, and the acquired replicas were measured using various precision instruments, including contact CMMs, optical projectors, toolmaker's microscopes, and image dimension measurement systems. The reference values were established using a contact CMM whose performance is verified according to the international standard ISO 10360-2. The experimental design consisted of two stages. In the first stage, a two-way ANOVA with replication was applied, with measurement equipment and slicing thickness as independent variables, to analyze the influence of each factor and derive optimal process conditions that minimize inter-device deviation. In the second stage, Two One-Sided Tests (TOST) for equivalence were conducted under the derived optimal conditions to verify that the indirect measurement values were statistically equivalent to the reference values. Additionally, a Gage R&R analysis was performed to assist in verifying that the repeatability and reproducibility of the system satisfied industrial tolerance criteria. The results demonstrated that the proposed PVS-based indirect measurement system exhibited statistically significant equivalence to the reference values for most complex geometric features, including the dovetail slots. In particular, the Gage R&R analysis confirmed that the variations within the measurement system were within a state of statistical control, thereby ensuring high measurement reliability. This study holds significant original value by elevating the PVS replication technique to a practical 'measurement system' through a systematic statistical validation process. These findings suggest the feasibility of the PVS-based indirect measurement system as a non-destructive and alternative measurement technology for precision parts with complex internal structures and are expected to serve as foundational data for enhancing quality control efficiency in various precision manufacturing fields.